Many studies have documented the value of multi-faceted interventions based on Wagner's Chronic Care Model for treating depression in primary care, but despite a well-known association between depression and intimate partner violence, (IPV) none have focused on the needs of depressed women with histories of IPV. Our long-range goal is to test the effectiveness of an abuse-sensitive chronic care model for depressed women seen in primary care settings. The objective of this proposal is to identify violence-related predictors of adherence and response to an existing multi-faceted depression care intervention, to adapt that model to be more responsive to the needs of abuse survivors, and to assess the feasibility and acceptability of the new abuse-sensitive model in preparation for a large randomized control trial. We hypothesize that women with a history of IPV are less likely to respond to traditional depression treatment models than non-abused women. We further hypothesize that the addition of special abuse-sensitive features such as abuse assessments, referrals to community resources, self-management tools about IPV, and group cognitive behavioral therapy focused on depression and abuse will improve the effectiveness of the chronic care model in treating depression. We propose 3 studies: 1) a prospective cohort study identifying violence related predictors of response to a traditional multi-faceted depression care intervention (care management, psychiatry support, provider education and a depression registry) being implemented by our collaborators; 2) a qualitative needs assessment of depressed IPV survivors to determine their beliefs, needs, treatment barriers and preferences; and 3) a pilot intervention assessing the acceptability and feasibility of an abuse sensitive depression care model. This work is significant, as it will improve our ability to provide mental health services to this challenging group of primary care patients. It is innovative as it applies concepts of chronic care management to an understudied and challenging population. This project will provide the necessary training, experience, and preliminary data for a R01 proposal to conduct a randomized controlled trial testing the effectiveness of an abuse-sensitive depression care model.